Revision Sinus Surgery

turbinate hypertrophy, or polypoidal disease, which may be a causative factor in the failure of DCR [8].

Surgical Technique for Revision Cases

If it is done under general anesthesia, hypotensive anesthesia is preferred. The inside of the nose is decongested. The head is elevated by bringing the table 20° or 30° in a reverse Trendelenburg position. The lateral nasal wall and previous rhinostomy site are examined carefully endoscopically for potential structural issues that may have contributed to the failure of initial procedure [38]. The mucosa anterior to the middle turbinate may be infiltrated with 1 ml lidocaine with 1:100,000 epinephrine. The disadvantage of this infiltration is that it may obscure the bulging of the nasolacrimal sac. It is important not to traumatize the nasal mucosa, since bleeding prevents visualization. At this stage any septal deviation, concha bullosa, or paradoxical middle turbinate obstructing the view

must be corrected. Occasionally, the head of the middle turbinate needs to be removed to expose the sac area. Chronic maxillary rhinossinusitis or pansinusitis should be addressed simultaneously by the same endonasal route. In revision cases, before starting the operation the lacrimal probe is introduced through the superior canaliculus and passed into the area of rhinostomy. In most cases a fibrous membrane is noted to occlude the previously created channel. The extent of the previous rhinostomy and the size of the bony opening are determined [39]. Active Wegener’s granulomatosis or benign or malignant lesions of the lacrimal sac or the neighboring structures are absolute contraindications for DCR (Table 27.8).

Localization of the Lacrimal Sac and Duct

The key initial landmark is the posterior border of the frontal process of the maxilla, which is usually identifiable as a ridge or an indentation into the nasal airway just

anterior to the middle turbinate [36]. This ridge extends from the highest point of the inferior turbinate upward and ends immediately in front of the middle turbinate attachment. The nasolacrimal duct and sac lie immediately lateral and posterior to this ridge. Superiorly, the duct joins the sac halfway between the attachments of the middle and inferior turbinates. The superior border of the lacrimal sac is above the middle-turbinate anterior attachment.

The average position of the apex of the lacrimal sac is 6.10 ± 2.02 mm (range, 2–12 mm) above the opercule of the middle turbinate [9]. The anterior attachment of the uncinate process is at the junction of the lacrimal fossa and the orbital plate of the lacrimal bone. The nasolacrimal sac is always situated immediately anterior to the uncinate process and makes the uncinate process a good landmark in DCR operations. It is important not to enter beyond this point in a lateral direction since it may cause orbital penetration. A 20-gauge fiberoptic light pipe may also be used to identify the location of the sac and is inserted into the lacrimal sac through either of the lacrimal canaliculi. The light is visualized endonasally with a rigid

as is done in the usual way. A rectangular cut is made in the mucosa anterior to the middle turbinate and superior to the inferior turbinate. Wong et al. [37] advise an oval cut 2 cm × 1 cm into the mucosa. Wormald et al. performs incision 3–5 mm posterior to and about 8–10 mm above the axilla and brings it 10 mm anterior to the axilla onto the frontal process of the maxilla. They extend the incision inferiorly about two-thirds of the length of the anterior end of the middle turbinate to the insertion of inferior turbinate.

Indeed, removal of nasal mucosa 7–8 mm in diameter might be enough and unnecessary removal mucosa should be avoided to decrease new scar formation. A canal knife, as used in ear surgery, can be used for this procedure. Since bleeding occurs mainly from the edge of the cut mucosa, it is important to make a complete cut, and not to pull and tear the mucosa [24]. After the cut is completed through the mucoperiosteum all the way down to the bone, the mucosa is elevated off the bone and removed or can be used as a posteriorly based flap. It does not appear to be necessary to apply special mucosal flaps [13]. If needed, through-cutting forceps may be used instead of Blakesley forceps to avoid tearing of the mucosa. To prevent recurrences it is important to remove the periosteum with the nasal mucosa and to remove all bony partitions and spicules. The periosteum induces new bone formation and in turn leads to narrowing or closure of the ostium. If the edge of the previous rhinostomy opening is found, the mucosa around it is elevated and the bone is drilled in such a way that the new rhinostomy be at the correct location. It is important to prepare the sac completely exposed.

Bone Removal

Inadequate bone removal is a common cause of failure in DCR. In all revision operations agger nasi cells should be checked. In approximately 8% of patients there is an agger nasi cell in this area. In some cases, agger nasi ethmoid cells may extend under the fossa causing confusion during surgery, and removal of bone results in opening into the ethmoid cell rather than the lacrimal fossa. If the ag-

Revision Dacryocystorhinostomy

ger nasi cell has not been opened yet, it will be necessary to open the agger nasi cell up and go through it before going through the lateral wall and the lacrimal bone into the sac.

If the anterior edge of the previous bony rhinostomy site is found, it is widened so as to create a proper rhinostomy. If there can be found no bony rhinostomy opening, the surgical procedure should be as in primary DCR operation. The sac and duct junction is above the insertion of inferior turbinate, 6 mm below the operculum. Woog et al. suggest removing the bone overlying the common canaliculus completely and achieving a rhinostomy with at least 6–8 mm vertical dimension in adults. The medial side of the bone of the maxillary portion of the lacrimal fossa can be removed either from posterior to anterior or from anterior to posterior. Since it is thinner in the posterior part, it makes sense to start from posterior. However, it is really challenging to remove the bone with conventional endoscopic sinus instruments alone. There are currently no bone-removing instruments specifically designed for this location. Kerrison forceps or backbiting forceps may be used for this purpose [40]. The surgeon may feel safer if he/she starts from the anteromedial part of the bony sac, identifies the sac, and continues posteriorly. The use of a laser takes more time and may cause thermal injury. The laser can only ablate the much thinner lacrimal bone. Removal of part of the frontal process of the maxilla gives better access and visualization of the lacrimal sac, but a laser cannot ablate this thick bone.

Weidenbecher et al. [33] suggest removing the entire medial bony covering of the sac. Whittet et al. [36] insert a Leibrich lacrimal probe into the inferior canaliculus, direct it against the medial wall of the lacrimal sac in order to tent, and decide how much bone to remove after this procedure. They advocate leaving approximately 5 mm free of bone around the canaliculus, especially at the junction of attachment of the middle turbinate and the lateral nasal wall, a point that demarcates the floor of the lacrimal fossa. Woog et al. [38, 39] suggest creating a generous rhinostomy that extends from above the middle-turbinate attachment to the level of the midpoint of the maxillary line inferiorly. Welham and Wulc [35] think that the ideal osteotomy should involve the removal of all of the bone between the medial wall of the sac and the nose. Drills specifically designed for intranasal use make it easier to remove the bone but may be associated with thermal injury and damage to the surrounding mucosa. The bony opening should start a few millimeters above the operculum, and the medial half of the bony nasolacrimal sac should be removed in such a way that after the surgical bony opening it should be 10–17 mm in diameter (average 11.75 mm). A functional result can be achieved with a fistula of 6 mm2. No statistical correlation

241

was found between the size of the ostium at the surgery and after healing [18]. Iliff [15] removed a 10-mm diameter piece of bone, with 1 failure out of 87 cases. The mean diameter of the healed ostium was 1.8 mm, representing an area only 18% of that of the initial anastomosis. So the aim is not to create a very large bony osteotomy, but a functioning osteotomy of efficient size.

The entire medial bony covering of the sac can be removed under endoscopic or microscopic control. If a chisel is used, attention should be given to removing all bony fragments, since if left they may cause obstruction later. Previously irradiated patients should be handled carefully owing to healing problems and anatomic derangements. However, in revision cases using drills is more advocated. Thus, following DCR, the sac and the duct should coexist as anatomic structures and be incorporated into the nose.

Removal of the Medial Mucosal Wall of the Sac

The vascularized white color of the sac is characteristic and can easily be identified. The medial part of the lacrimal sac should be fully exposed. A lacrimal probe may be used to identify the lacrimal sac. A lacrimal probe is passed through a canaliculus and directed medially into the obstructed sac. The fibrous tissue is tented into the nasal cavity with probes to provide a broad soft-tissue region under tension. The tenting of the medial sac wall by the probe is visualized endonasally. While it is tented by the lacrimal probe, the sac mucosa is incised with a sickle knife. Mucopus, residual contrast material, may drain from the sac, or dacryoliths may be seen in the interior of the sac. Once the sac has been entered, the lacrimal probe may be seen. Using Bellucci scissors, one can extend the incision and use through-cutting instruments to enlarge the intranasal opening. Blakesley forceps are not through-cutting and may tear the sac mucosa; for this reason, through-cutting forceps are preferred. A carbon dioxide laser can also be used. As much of the medial wall of the sac should be removed as possible. Metson [19] advises enlarging of intranasal opening to a diameter of approximately 10 mm, allowing free passage of the lacrimal probes into the nose from both superior and inferior canaliculi. In some failed DCRs there is a large sac remnant seen in DCG whereas in others a small cicatrized sac is present. In patients with small cicatrized sacs it may be difficult to achieve lacrimal and nasal mucosal anastomosis. If there exists any problem to appose the mucosal edges, the nasal mucosal flap is trimmed to fit the size of the lacrimal ostium [28]. According to Woog et al. [39] an adequate rhinostomy should permit easy passage of lacrimal probes and removal of the medial wall of the

242

lacrimal sac in the area of common canaliculus should be confirmed by direct visualization of the internal common punctum with angled endoscopes.

In revision procedures orbital soft tissues and internal common punctum may be incorporated into the scar tissue. Vigorous avulsion of this cicatrix may cause injury to orbital soft tissues, medial canthal ligament, and internal common punctum. Woog et al. [38, 39] recommend to observe the medial commissure while gentle traction is placed on the tissue to be removed at the rhinostomy site. Excessive movement of the medial commissure by this maneuver may signify that deeper tissues than desired are being grasped by the forceps.

Silicone Tubing

A silicone tube is placed through the upper and lower canaliculi into the nasal cavity, the ends of the tubing are grasped with forceps, guided out of the nose, and are tied and trimmed so that the knot lies within the nasal cavity. The tubing thus forms a continuous loop, which passes through the intranasal ostium and is unlikely to become dislodged [19]. The knot may be fixed by a suture or a vessel clip. Wong et al. [37] use a black silk suture to tie round the silicone tubes. According to Allen and Berlin [1] and Bartley [2], silk sutures can produce pyogenic and giant cell granulomas. Packing of the nose is unnecessary unless bleeding is a problem.

Silicone intubation may be recommended in cases with canalicular stenosis, a small scarred lacrimal sac, a tight upper nasal cavity, in reoperations, and if the flaps of the lacrimal sac and nasal mucosa are not sutured. In other cases it is the surgeon’s preference to use silicone tubes or not. Silicone tubing may serve to dilate constricted passages in patients with canalicular or common internal punctual stenosis and marks the site of the intranasal ostium [6]. Kohn [17] believes that silicone tubing keeps the anterior and posterior flaps separate, and also discourages cicatricial closure of the bony ostium. Some reports [1, 23] give less favorable results when tubes are inserted. Allen and Berlin [1] believe that intubation may be the reason for failure. Tubes may cause inflammation, granulation, and slit the canaliculus. Silicone tubing may incite granulomatous inflammation at the internal ostium, chronic infection, or canalicular laceration. Walland and Rose [30] did not find any difference in failure rates or inflammation with or without silicone intubation. Insertion of silicone tubes is generally encouraged [3]. Snead et al. [26] showed that silicone intubation did not increase canalicular inflammation in animals. Walland and Rose [31] found no significant difference in the

27 rate of failure or soft-tissue infection between silicone-in- tubated and nonintubated patients. Wormald et al. placed

Metin Onerci

tubes to dilate the common canaliculus opening into the lacrimal sac; this is especially important for patients who have functional epiphora rather than to keep the sac open. In revision cases intubation with silicone tubes is generally recommended.

How Long Should the Silicone Tubes Be Kept in Place?

It is advised to keep silicone stents in place for 2–6 months; however, tubes kept in place over 3 months are associated with inflammation and granulation. Wong et al. [37] and Weidenbecher et al. [33] advise removing the tubes after 6 weeks. El Guindy et al. [8] keep the tubes for only 2 months and recommend early removal (at 2 months) because silicone tubing may incite a granulomatous reaction. Hartikainen et al. [10] keep the tubes in place for 6 months. Hausler and Caversaccio [11] use the tubes in place for the long term and they have several cases where silicone tubes have been used for over 3 years without any complications. They even suggest leaving the tubes permanently in persistent cases and think that the silicone tubing produces a maximal dilatation of the canaliculi and a natural aspiration of tear liquid by capillary force. However, keeping silicone tubes for at least 2 months is generally advocated in revision cases.

Mitomycin C

Mitomycin C, an antiproliferative agent that is widely used in pterygium excision and trabulectomy with favorable results, was also used to inhibit fibrous tissue growth and scarring at the osteotomy site and to decrease the failure rate. Cottonoids soaked with 0.2 mg/ml mitomycin C are applied to the osteotomy site. Kao et al. [16] reported that mytomicin C improved success rates; Zilelioğlu et al. [42] found no benefit in using mitomycin C. Use of mitomycin C needs further investigation.

Complications

Soft-tissue infection after open lacrimal surgery occurs in 8% of patients. It is reduced fivefold with routine administration of antibiotics [31]. Vardy and Rose [29] demonstrated that intraoperative or postoperative broad-spec- trum antibiotics reduced the incidence of cellulitis after open primary lacrimal surgery. Tsirbas and McNab [27] reported 3.8% of cases with secondary hemorrhage after DCR. The incidence of bleeding was higher in patients taking nonsteroidal anti-inflammatory drugs. Severe nasal hemorrhage requiring nasal packing was also reported

Revision Dacryocystorhinostomy

[22, 34]. Orbital complications were also mentioned [25, 32]. Orbital emphysema can be prevented by asking patients not to blow their nose for at least 2 weeks following surgery. Prolapsed tubes, punctal widening, corneal irritation, and intranasal discomfort are the complications related to silicone tubing [5]. There are rare reports of cerebrospinal fluid leaks and meningitis [4, 12, 20]. Meticulous surgery avoiding unnecessary mucosal trauma may prevent intranasal adhesions.

Tips and Pearls

1.The sac should not be opened very high up without opening it inferiorly in order to prevent sump syndrome. High and small rhinostomies may cause sump syndrome.

2.Preservation of as much mucosa as possible is of paramount importance to decrease scarring.

3.The bony osteotomy should be as big as the mucosal opening (7–10 mm) and the periosteum should be removed with mucosa (since new bone formation requires the presence of periosteum).

4.The bone in the region of the lacrimal fossa may be thicker following a midfacial fracture owing to callus formation.

5.The medial membranous sac wall should be removed. Only puncturing the medial membranous wall decreases the success rate.

Conclusion

A small bony ostium, inadequate rhinostomy, excessive scar tissue production, anatomical abnormalities, and concomitant paranasal sinus infections are the most important causes of the failure of DCR surgery. The endonasal approach is well suited for revision DCR surgery because the residual lacrimal sac can be accessed directly through the previous bony ostium created at the primary DCR. The bony ostium is more easily enlarged through the endonasal approach. It preserves the pump function of the sac and avoids an external scar. It allows sinus surgery to be performed in the same sitting if needed. The success of the surgery depends on accurate diagnosis ruling out canalicular and functional obstruction, and meticulous surgery, avoiding trauma to the mucosa and neighboring structures.

References

243

2.Bartley GB (1992) Acquired lacrimal drainage obstruction: an etiologic classification system, case reports, and a review of the literature. Part 1. Ophthal Plast Reconstr Surg 8:237–242

3.Beigi B, Westlake W, Chang B, Marsh C, Jacob J, Chatfield J (1998) Dacryocystorhinostomy in South West England. Eye 12:358–362

4.Beiran I, Pikkel J, Gilboa M, Miller B (1994) Meningitis as a complication of dacryocystorhinostomy. Br J Ophthalmol 78:417–418

5.Brookes JL, Oliver JM (1999) Endoscopic endonasal management of prolapsed silicone tubes after dacryocystorhinostomy. Ophthalmology 106:2101–2105

6.Burns JA, Cahill KV (1985) Modified Kinosian dacryocystorhinostomy. Ophthalmic Surg, 16:710–716

7.Dutton JJ (1988) Diagnostic tests and imaging techniques. In: Linberg JV (Ed) Lacrimal Surgery. Churchill Livingstone, New York, pp 19–48

8.El-Guindy A, Dorgham A, Ghoraba M (2000) Endoscopic revision surgery for recurrent epiphora occurring after external dacryocystorhinostomy. Ann Otol Rhinol Laryngol 109:425–430

9.Fayet B, Racy E, Assouline M, Zerbib M (2005) Surgical anatomy of the lacrimal fossa. A prospective computed tomodensitometry scan analysis. Ophthalmology 112:1119–1128

10.Hartikainen J, Antila J, Varpula M, Puukka P, Seppa H, Grenman R (1998) Prospective randomized comparison of endonasal endoscopic dacryocystorhinostomy and external dacryocystorhinostomy. Laryngoscope 108:1861–1866

11.Hauslaer R, Caversaccio M (1998) Microsurgical endonasal dacryocystorhinostomy with long term insertion of bicanalicular silicone tubes. Arch Otolaryngol Head Neck Surg 124:188–191

12.Heerman J Jr (1991) Rhinologische Aspekte bei Traenenwegstenosen. Otorhinolaryngol Nova 1:227–232

13.Hosemann WG, Weber RK, Keerl RE, Lund VJ (2002) Minimally Invasive Endonasal Sinus Surgery. Georg Thieme Verlag, Stuttgart, pp 66–70

14.Hurwitz JJ (1996) The Lacrimal System. Lippincott Raven, Philadelphia

15.Iliff CE(1971) A simplified dacryocystorhinostomy. Arch Ophthalmol 85:586–591

16.Kao SCS, Liao CL, Tseng JHS, Chen MS, Hou PK (1996) Dacryocystorhinostomy with intraoperative mitomycin C. Ophthalmology 104:86–91

17.Kohn R (1988) Ophthalmic Plastic and Reconstructive Surgery. Lea Febiger, Philadelphia

18.Lindberg JV, Anderson RL, Bumsted RM, Barreras R (1982) Study of intranasal ostium after external dacryocystorhinostomy. Arch Ophthalmol 100:1758–1762

244

20.Neuhaus RW, Baylis HI (1983) Cerebrospinal fluid leakage after dacryocystorhinostomy. Ophthalmology 90:1091–1095

21.Onerci M (2002) Dacryocystorhinostomy. Rhinology 40:49–65

22.Orcutt JC, Hillel A, Weymuller EA Jr (1990) Endoscopic repair of failed dacryocystorhinostomy. Ophthal Plast Reconstr Surg 6:197–202

23.Psilas K, Eftaxias V, Kastaniondakis J, Kalogeropoulos C (1993) Silicon intubation as an alternative to dacryocystorhinostomy for nasolacrimal drainage obstruction in adults. Eur J Ophthalmol 3:71–76

24.Shun-Shin GA (1998) Endoscopic dacryocystorhinostomy. Eye 12:467–470

25.Slonim CB, Older JJ, Jones PL (1984) Orbital hemorrhage with proptosis following a dacryocystorhinostomy. Ophthalmic Surg 15:774–775

26.Snead JW, Rathbun JE, Crawford JB (1980) Effects of silicone tube on the canaliculus. An animal experiment. Ophthalmology 87:1031–1036

27.Tsirbas A, McNab AA (2000) Secondary hemorrhage after dacryocystorhinostomy. Clin Exp Ophthalmol 28:22–25

28.Tsirbas A, Davis G, Wormald PJ (2005) Revision dacryocystorhinostomy, Am J Rhinol 19:322–325

29.Vardy SJ, Rose GE (1999) Prevention of cellulitis after open lacrimal surgery. Ophthalmology 107:315–317

30.Walland MJ, Rose GE (1994) The effect of silicone intubation on failure and infection rates after dacryocystorhinostomy. Ophthal Surg 25:597–600

31.Walland MJ, Rose GE (1994) Soft tissue infections after open lacrimal surgery. Ophthalmology 101:608–611

Metin Onerci

32.Weber R, Draf W, Kolb P (1993) Die endonasale mikrochirurgische Behandlung von Traenenwegsstenosen. HNO 41:11–18

33.Weidenbecher M, Hosemann W, Buhr W (1994) Endoscopic endonasal dacryocystorhinostomy. Ann Otol Rhinol Laryngol 103:363–367

34.Welham RA, Hughes SM (1985) Lacrimal surgery in children. Am J Ophthalmol 99:27–34

35.Welham RAN, Wulc AE (1987) Management of unsuccessful lacrimal surgery. Br J Ophthalmol 71:152–157

36.Whittet HB, Shun-Shin GA, Awdry P (1993) Functional endoscopic transnasal dacryocystorhinostomy. Eye 7:545–549

37.Wong RJ, Glicklich RE, Rubin PAD, Goodman M (1998) Bilateral nasolacrimal duct obstruction managed with endoscopic techniques. Arch Otolaryngol Head Neck Surg 124:703–706

38.Woog JJ, Sindwani R(2006) Endoscopic dacryocystorhinostomy and conjuctivodacryocystorhinostomy. Otolaryngol Clin N Am 39:1001–1017

39.Woog JJ, Kennedy RH, Custer PL, Kaltreider SA, Meyer DR, Camara JG (2001) Endonasal dacryocystorhinostomy: a report by the American Academy of Ophthalmology. Ophthalmology 108:2369–2377

40.Yung MW, Logan BM (1999) The anatomy of the lacrimal bone at the lacrimal bone at the lateral wall of the nose. Clin Otolaryngol 24:262–265

41.Wormald PJ (2006) Powered endoscopic dacryocystorhinostomy. Otolaryngol Clin North Am 39:539–549

42.Zilelioğlu G, Uğurbaş SH, Anadolu Y, Akıner M, Aktürk T (1998) Adjunctive use of mytomicin C on endoscopic lacrimal surgery. Br J Ophthalmol 82:63–66

■A multidisciplinary team should be utilized including an endocrinologist and neuro-ophthalmologist, in addition to the neurosurgeon and otolaryngologist who will be performing the procedure, to maximize perioperative treatment.

■Computed tomography and magnetic resonance imaging should be used as complementary imaging studies.

■Intraoperative stereotactic image guidance is crucial for safe revision endoscopic approaches to the sella.

■Revision endoscopic transsphenoidal hypophysectomy is associated with an increased risk of cerebrospinal fluid leak irrespective of the prior method of resection.

Introduction

Approaches to the sella have varied greatly over the past 100 years, with Schloffer performing the first transsphenoidal approach to the pituitary in 1907. This was later modified by Cushing in 1914 when he described the sublabial transseptal approach to the sella, but he eventually abandoned this approach secondary to concern for high recurrence rates. With the advent of fluoroscopy and the operating microscope, the sublabial transseptal approach again came into favor in the 1970s, decreasing the overall morbidity associated with accessing the sella via a frontal craniotomy [3]. The next major advancement came in 1992 when Jankowski et al. first reported the successful use of 0 and 30 endoscopes to approach the sella via a transnasal approach, and since that time numerous reports have outlined the surgical technique and overall safety of the procedure [1, 5, 6, 12]. The aim of this chapter is to review revision endoscopic transsphenoidal

hypophysectomy, focusing specifically on the indications and preoperative considerations, surgical technique, and complications.

Indications for Surgery

The indications for revision pituitary surgery are similar to those for primary surgery. Recurrence of pituitary lesions can occur months to years following primary resection. Surgical intervention is generally reserved for:

1.Pituitary adenomas that are nonresponsive to medical management.

2.Pituitary masses that cause visual impairment.

3.Lesions that are increasing in size, as evidenced with serial imaging.

4.Pituitary apoplexy.

Pituitary apoplexy is considered a surgical emergency and occurs when necrosis or hemorrhage into pituitary lesions causes abrupt vision loss, headache, cranial neuropathies, and sometimes acute adrenal insufficiency.

Exclusively suprasellar tumor. Relative contraindications are:

1.Prior craniotomy.

2.Multiply recurrent tumor.

Preoperative Evaluation

Prior to transsphenoidal hypophysectomy a patient should be evaluated by a multidisciplinary team to ensure proper management in the perioperative period. In addition to the neurosurgeon and otolaryngologist who will be performing the procedure, this team should include an endocrinologist and neuro-ophthalmologist.

■Preoperative anesthesia consultation should be considered, especially in those patients with acromegaly and Cushing’s disease.

Excessive growth hormone affects the heart, leading to cardiac myopathy, as well as pharyngeal tissue, leading to hypertrophy of the base of tongue and redundant mucosa, creating a potentially difficult intubation, and may require tracheotomy. Patients with Cushing’s disease have increased intraoperative anesthetic risk due to their multiple comorbidities including obesity, diabetes, obstructive sleep apnea, and hypertension.

As mentioned, patients with acromegaly may require a temporary tracheotomy to establish a safe airway, and this should also be considered in patients with severe obstructive sleep apnea, as these patients will not be able to use continuous positive airway pressure (CPAP) postoperatively. In addition, the negative pressure from deep breaths associated with an apneic event can lead to pneumocephalus postoperatively. Similarly, wide swings in intracranial pressure occurring during apneic events may increase the risk of postoperative cerebrospinal fluid (CSF) leak.

A detailed history and physical examination should be performed. Special care should be taken to note any history of sinonasal complaints including nasal congestion or obstruction, hyposmia/anosmia, rhinorrhea, postnasal drip, past nasal trauma, and headache. In addition, history of other medical problems that could impact the surgical outcome should be specifically queried including diabetes, hypertension, and obstructive sleep apnea. Review of old operative reports should be performed, noting the type of approach (sublabial transseptal, frontal craniotomy, or endoscopic endonasal), side of approach when applicable, and any perioperative complications. A

Karen A. Kölln and Brent A. Senior

detailed physical examination should be completed specifically to evaluate for visual field deficits, visual acuity, stigmata of endocrine dysfunction, and cranial neuropathies. Sinonasal endoscopy should be performed in all patients preoperatively to evaluate for septal deviation, nasal polyposis, and purulence.

■Imaging is of utmost importance in preoperative, as well as intraoperative, surgical planning in revision surgery, as scarring can lead to distortion of the anatomy including medialization of the carotid arteries.

Imaging should include the complementary studies of a noncontrasted computed tomography (CT) of the paranasal sinuses with fine (< 3-mm) cuts and a magnetic resonance imaging (MRI) scan of the brain with gadolinium enhancement, including special focus on the pituitary gland. CT of the paranasal sinuses is used to better define the bony anatomy, pneumatization of the sphenoid, presence of a concha bullosa, position and possible dehiscence of the carotid artery, location of the intersinus septum, presence of an Onodi sinus, and extent of postsurgical scarring (Fig. 28.1). MRI of the brain should be used to define the extent and location of the tumor, including cavernous sinus involvement and impingement of the optic nerve/chiasm, and very importantly, to assess for any vascular abnormalities and the location of the carotid arteries (Fig. 28.2). These images should then be fused for intraoperative stereotactic guidance.

Fig. 28.1  Computed tomography, sagittal view, demonstrating previous resection and scar within the sphenoid

Laboratory data including complete blood count, and a comprehensive metabolic panel to evaluate for hyponatremia, hypokalemia, hypercalcemia and hyperglycemia as well as other metabolic abnormalities should be obtained. Coagulation studies including partial thromboplastin time, prothrombin time, and international normalized ratio should be obtained if there is a family history of bleeding disorders. Endocrine evaluation of prolactin, insulin-like growth factor-1, adrenocorticotropic hormone, thyroid-stimulating hormone, thyroxine, follicle-stimulating hormone, luteinizing hormone, testosterone, morning cortisol, 24-h urine free cortisol, and 24-h urine free cortisol with low-dose dexamethasone suppression should occur as indicated by the clinical scenario [8]. In addition, all of our patients are placed on perioperative steroids due to decreased endogenous steroid production.

Surgical Technique

Standard-length 4-mm endoscopes fitted with scope irrigation are used to approach the sella, as they afford the

best visualization and illumination. Angled (45 and occasionally 70) and longer telescopes can be useful in tumor resection and exploration of the sella. When approaching the sella, the surgeon can use a unilateral or bilateral approach. Early in our experience we preferred the unilateral approach, with the side of the approach determined primarily by nasal factors: the presence of septal deviation and concha bullosa, the degree of nasal congestion, and the location of the sphenoid intersinus septum. If all nasal factors are considered equal, the contralateral side usually confers the better angle of approach in lesions that are off the midline or that extend into the cavernous sinus.

■We have found with time that the bilateral approach is usually preferred as this allows wide access to the sella and the ability to use one side for the endoscope while the other is used for instrumentation. In addition, this can be requisite in patients with thin nasal cavities.

The patient is positioned in the “beach-chair” position with the head and torso elevated and the knees bent, and the patient’s right arm is tucked by their side. The patient’s head rests in a donut (gel or foam) and the head is rotated

slightly toward the surgeon in a patient with a large body habitus. Greater palatine blocks are then performed to aid in hemostasis by injecting 1.5 ml of 1% lidocaine with 1:100,000 epinephrine transorally into each greater-pala- tine canal. The patient is then registered with the stereotactic computer guidance software to ensure safe tumor removal. The abdomen is prepped and draped in a sterile manner in case a fat graft is needed; however, the face is not prepped as the instruments are passed through a contaminated nasal cavity. Under endoscopic guidance, pledgets soaked with 0.05% oxymetazoline hydrochloride are placed into each nasal cavity for decongestion and then additional lidocaine is injected at the junction of the horizontal portion of the basal lamella and the lateral nasal wall to achieve a sphenopalatine artery block. The sphenoid ostium is then identified; a variable amount of scar tissue may be present both intranasally and at the face of the sphenoid.

■ We rely heavily on image guidance to aid in the identification of the sphenoid, which can be entirely scarred over.

If scar has formed between the middle turbinate and septum, it is lysed with a cutting instrument. Stapes and J-curettes are utilized initially, and then the opening is serially enlarged using a mushroom punch or Kerrison rongeur. In some cases the sphenoid had been previously packed with fat and this must be carefully cauterized and resected using cutting instruments until the face of the sella is reached (Video 28.1).

■ Great care must be taken when resecting tissue within the sphenoid, as the carotid and optic nerves can be dehiscent in 10 and 4% of individuals, respectively [11].

Once the sella has been adequately exposed, the mucosa on the posterior wall of the sphenoid sinus is coagulated with a bipolar cautery. Knowledge of the previous resection and presence of a CSF leak can aid the surgeon at this point for safe entry into the sella. There can be significant bone regrowth and it may be necessary to use a highspeed drill, or else a Kerrison may simply be needed to enlarge the previous opening into the sella. The edges of the window into the sella should be adequately exposed in any case, as demonstrated in Video 28.2. The endoscope is then attached to a fixed pneumatic holder, or an assistant can hold the scope, allowing the operating surgeon both hands for instrumentation. The dura is subsequently cauterized with bipolar cautery and a sickle knife is used to make a cruciate incision. This step again can

Karen A. Kölln and Brent A. Senior

be encountered with varying difficulty, as the dura may be scarred and it can be difficult to incise into the dura without causing a CSF leak. We have found that revision surgery is associated with an increased risk of CSF leak, regardless of the prior method of resection.

■In revision hypophysectomy the carotid arteries can be tethered medially; at this point the fused MRI and CT images are crucial for safe entrance into the sella and subsequent tumor removal.

The tumor generally bulges through the opening and samples are sent for frozen section. The remaining tumor is removed using neurosurgical ring curettes and suction (Video 28.3). In order to ensure complete tumor removal, the endoscope can be inserted inside the sella and angled telescopes can be used to examine the lateral crevices of the sella. The technique of “hydroscopy,” the system of normal saline irrigation under pressure flooding the sella, can be utilized to push the diaphragma up, as well as to wash away debris and clot [13].

Following complete tumor removal, hemostasis is obtained by placing a hemostatic substance, such as microfibrillar collagen in thrombin, over the operative field. When this is washed away the patient is evaluated for a CSF leak.

■If no CSF leak is encountered we have found that reconstruction of the sella is not necessary.

In our report in 2003 we determined that there was no increase in the rate of postoperative leak (0.4% compared to reported rates of 0.8–6.4%) and there were no cases of meningitis or empty-sella syndrome [15]. If a CSF leak is encountered, we reconstruct the sella by using microfibrillar collagen and a fat graft bolstered by an absorbable miniplate cut to fit under the edges of the sella.

Tips to Avoid Complications

1.Meticulous dissection should be performed, as incomplete resection is associated with an increased risk of hemorrhage into the tumor.

2.Consider temporary tracheotomy in patients with obstructive sleep apnea to minimize the risk of postoperative pneumocephalus and reduce the risk of postoperative CSF leak. In addition, consideration for tracheotomy should be made since CPAP will not be able to be utilized during the postoperative period.

3.Perioperative steroids are critical in all patients, given the disruption in endogenous steroid production.